Review of the Adjuvant Treatment of Breast Cancer: ASCO 2007

Review of the Adjuvant Treatment of Breast Cancer: ASCO 2007


The 2007 American Society of Clinical Oncology (ASCO) meeting featured a large number of breast cancer presentations addressing a broad array of topics. In the adjuvant arena, studies focused on advances in basic biologic understanding of different subtypes of breast cancer and the way this knowledge can translate into better, rationally designed therapies. At the same time, some of our long-held assumptions were challenged. As always, ASCO 2007 reports enhanced our understanding of breast cancer while raising a number of important questions.

Adjuvant Chemotherapy

It is generally accepted that three drug regimens, including an anthracycline, cyclophosphamide and a taxane provide the most benefit in patients with node-positive breast cancer. However, the optimal dosing and sequencing of these drugs is controversial.

When the ECOG 1199 trial was designed in the late 1990s, there was great debate about paclitaxel versus docetaxel’s activity in breast cancer. Equally controversial was whether weekly treatment was superior to standard every three-week (Q 3 week) dosing. E1199 randomized patients to receive doxorubicin and cyclophosphamide (AC) followed by either weekly or every three-week paclitaxel versus docetaxel in a 2 x 2 study design. About 5,000 women with lymph node-negative or positive disease, who were HER2-or-HER2+ were randomized. The report presented by Sparano, et al on behalf of ECOG provided an update of the trial at a median follow up of five years.

In terms of overall dose delivery, 95% of every-three-week paclitaxel was actually administered in this study, while 75% of weekly docetaxel doses were delivered. Severe toxicity, particularly hematologic, was reported most in the Q 3 week docetaxel arm. Weekly dosing yielded less hematologic toxicity but more neuropathy with both taxanes. For the primary comparison of disease free survival (DFS), there was no difference between paclitaxel versus docetaxel or for every-three-week dosing versus weekly and therefore the study was a negative one.

Secondary comparisons demonstrated a higher five-year DFS rate for weekly paclitaxel, 81.5%, and Q 3 week docetaxel, 81.2%, compared to the other two arms with an odds ratio (OR) of 1.27 and 1.23, respectively. Weekly paclitaxel had better overall survival, OR 1.32, p=0.01, compared to the other three arms. Overall, while the primary endpoint showed no significant findings, the unexpected interaction between agent and schedule favored the weekly paclitaxel and the Q 3 week docetaxel approaches.

Continuing the theme of taxane scheduling was the study presented by Loesch, et al, a Phase III trial comparing doxorubicin and cyclophosphamide followed by Q 3 week paclitaxel (AC-P) to doxorubicin and Q 3 week paclitaxel followed by weekly paclitaxel (AP-P). This 1,830 patient trial of high-risk, lymph node-negative (LN-) and lymph node-positive (LN+) patients had a high percentage of HER2+ entrants, 34% overall. Nonetheless, the overall results were good in both arms: 85% DFS for AC-P versus 88% for AP-P, p=0.05, HR=0.74. The five year DFS demonstrated no difference in outcome with an HR=0.94, p=0.31. However, the overall survival was better by an absolute value of 3% at five years, 90% versus 87%, HR=0.76 for the weekly paclitaxel arm due to fewer breast cancer deaths.

Interestingly, the five year disease-free survival was worse for triple-negative patients compared to HER2+ patients in the Q 3 week paclitaxel arm despite the absence of HER2-directed therapy in this trial. Sensory neuropathy was worse in the weekly paclitaxel arm, perhaps because of the cumulative affect of multiple doses of paclitaxel through the treatment.

Taken together, these two studies support the use of weekly paclitaxel as a superior strategy in the adjuvant setting. The Loesch study does not really answer the question of the necessity of cyclophosphamide in adjuvant treatment as the preservation of DFS in the non-cyclophosphamide arm may have been attributable completely to the weekly paclitaxel. Although overall survival was superior, the lack of correlation with DFS is concerning, and toxicity was increased. Therefore, to my mind, the best arm of E1199, AC x 4 followed by weekly paclitaxel x 12 doses, seems to offer the highest benefit to risk ratio with high activity and low toxicity. However, in patients with a contraindication to anthracyclines, the AP followed by weekly paclitaxel arm represents an excellent new choice.

New information was provided regarding the interaction of hormonal status with chemotherapy. There is a growing body of evidence suggesting adjuvant chemotherapy is less effective in patients with ER-positive disease compared to ER-negative. A previously published study retrospectively addressing sequential CALGB trials utilizing paclitaxel after AC demonstrated this phenomenon clearly. At ASCO 2007, a study presented by Andre, et al, evaluated two previously reported trials utilizing docetaxel for the effect of ER status on outcome. Seventy-five percent of patients were ER+. The HR for death was similar in ER- and ER+ patients. The five-year absolute survival benefit was 7% for ER- and 4% for ER+. There was no significant interaction between docetaxel efficacy and ER status. These data support a potential difference between taxanes as docetaxel benefits both ER+ and ER- patients while the results with paclitaxel suggest benefit almost entirely confined to ER- status. It would be of great interest to perform a similar analysis in E1199 to provide more information about this important question.

HER2 Positive Disease

The initial presentation of the combined analysis of B-31 and N-9831, two randomized trials that added trastuzumab to chemotherapy in HER2+ patients, was a practice-changing experience. In these studies, patients were randomized to receive sequential AC followed by paclitaxel in the control arms (AC-T) or the same chemotherapy with trastuzumab added with the paclitaxel and continued for a total period of 52 weeks (AC-TH). Trastuzumab yielded an early and dramatic benefit with approximately a 50% improvement in disease-free survival. The results at the ASCO 2005 meeting, representing information based on two year data, were preliminary. Therefore the follow up on these results reported this year by Edith Perez were eagerly awaited.

Additional aspects of the trials were presented for the first time. For instance, an interesting finding was that overall 20.9% of the patients randomized to the non-trastuzumab arms actually received trastuzumab during study. However, the analysis presented both previously and updated at ASCO 2007 was an intent-to-treat analysis, and therefore probably underestimates the benefit of trastuzumab compared to non-trastuzumab regimens.

At four years of follow up, the disease-free survival for doxorubicin and cyclophosphamide followed by paclitaxel and trastuzumab (AC-TH) was 85.9% versus 73.1% for AC-T, HR=0.41. The event rates are highest in years two and three but continued to be significant in years four and five suggesting that these patients remain at risk for a substantial period of time. All subgroups benefited by the addition of trastuzumab. Overall survival was significantly better for AC-TH at 92.6% at four years versus 89.4% for AC-T, HR=0.65, p=0.007. The most significant toxicity was congestive heart failure which occurred in only 22 of 875 patients receiving AC-TH. The remarkable improvements in HER2+ patients with the addition of trastuzumab has, therefore, been maintained with longer follow up with these patients.

Central FISH and IHC testing was done on all patients in this study. Not all patients who had been locally believed to be HER2+ turned out to be so on central testing. Results of disease-free survival benefit in these subgroups were certainly provocative. While those patients with central IHC 3 + and FISH > 2.0 ratio (the definitions for positivity) had the most benefit with an HR of 0.47, a group of 218 patients were FISH+ and IHC negative. Results indicate that these patients derived no benefit from trastuzumab. Conversely, the small group of IHC 3+ and FISH- patients did derive benefit, HR=0.61 and those who were IHC 0-2+ and FISH- by central testing had an HR= 0.51 similar to the large cohort of centrally defined clear-cut positives.

What is the explanation of trastuzumab’s effect on patient groups that were traditionally felt not to benefit from the antibody? Paik, et al. reviewed the data from B-31 and elegantly demonstrated that we still have much to learn about what constitutes HER2 positivity and responsiveness to HER2 directed therapies. In B-31, patients were required to be IHC+ and/or FISH+ by a local determination in any lab. During the course of the trial, an amendment designated those labs that would be acceptable for testing after heterogeneity in test results became obvious. Subsequently as noted above, all samples were retested for IHC and FISH in a central laboratory. Both IHC and FISH were negative in 9.7% of patients by the central assay. However, in B-31 similar to the combined analysis presented by Perez, all subsets benefited from trastuzumab: FISH+, FISH-, IHC 3+, IHC <3+ and FISH- and IHC <3+.

In an attempt to explain these paradoxical results, Paik reviewed preclinical data and clinical data which convincingly illustrated the fact that HER2 expression as measured at the mRNA level or the protein expression level, is really a continuous parameter rather than a dichotomous one. It has long been known that in vitro, trastuzumab enhances chemotherapy in HER2 IHC 1-2+. In vivo, patients with FISH negativity do derive some benefit from trastuzumab added to chemotherapy, although certainly less than FISH+ patients. So perhaps we have to redefine what actually HER2 negativity is.

The heterogeneity of breast cancer supports the concept that some individual cancer cells may have amplification of HER2 while others have normal HER2 gene copy numbers which leads to an aggregated FISH score reported in the range of 1-2.0, currently considered negative. This is not an insignificant number as up to 25% of patients fall into this category. Paik recommends a clinical trial investigating the effect of chemotherapy and trastuzumab in this group because it is a significant proportion with a potentially large benefit from the addition of anti-HER2 therapy.

The downside of trastuzumab usage is well known – cardiac dysfunction. Rastogi, et al, provided a five-year update of cardiac dysfunction in B-31. In this study, careful screenings with a pretreatment MUGA and follow up exams at quarterly intervals were obtained. Overall, 14% of patients discontinued trastuzumab therapy due to changes in MUGA scans. A careful analysis of cardiac events through the course of the trial was conducted. Gratifyingly, now with five years of follow up, the cumulative incidence of cardiac events from trastuzumab remained essentially unchanged: 3.8% of patients in the trastuzumab arm had a cardiac event compared to 0.9% in the AC-T arm. Of patients confirmed with congestive heart failure (CHF), only 5/33 had symptoms during their last follow up. The most important risk factors for CHF were age >50, hypertensive medications, a baseline ejection fraction <54% and a post-AC ejection fraction of <54%.

These now mature efficacy and toxicity results from the trastuzumab studies make us more confident of recommending trastuzumab-based regimens to our patients when HER2 is overexpressed. The next generation of questions has already been raised – who really benefits from trastuzumab? Should we be using a small molecule HER2 inhibitor like lapatinib in place of, or in addition to trastuzumab in HER2+ patients? How long do patients need trastuzumab therapy in the adjuvant setting? Future ASCO meetings will undoubtedly answer these questions.


Given the great prognostic and predictive value of HER2, other markers of therapy are actively being explored. In an analysis of a study comparing TAC versus FAC, Dumontet, et al. evaluated tumor blocks for various potential biomarkers. Both p53 expression and expression of the protein Tau led to a worse outcome. Further studies to validate the potential impact of measuring these proteins by IHC are ongoing.

There has been a good deal of interest in assessing topoisomerase 2A expression, an enzyme which is the known target for anthracyclines. Prior work has fairly convincingly demonstrated that topo 2A amplification or overexpression is associated with a significantly improved response to anthracycline therapy. Both topoisomerase 2A and the HER2 gene are located near one another on the long arm of chromosome 17 and are sometimes co-overexpressed as part of an amplicon consisting of multiple genes in the same region. Press, et al. explored topo 2A expression in multiple data sets including the original pivotal metastatic trastuzumab trial, and BCIRG 006 which randomized patients to AC-T, AC-TH, or carboplatinum/docetaxel/trastuzumab (TCH). In the absence of trastuzumab, coamplification of topo 2A and HER2 was associated with a longer DFS. However, in the trastuzumab arms, no difference in DFS with topo 2 amplification was noted. Moreover, addition of trastuzumab to anthracycline-based therapy does not appear to prolong DFS in the presence of topo 2A and HER2 coamplification.

These interesting results suggest that further individualization of therapy for breast cancer should occur, allowing us to move away from the “one size fits all” approach to treatment that categorized the last 20 years of adjuvant chemotherapy. As we now utilize HER2 and ER/PR testing to determine best therapy, integration of these emerging biomarkers into clinical decision making may allow us to better predict relapse and offer improved therapeutic approaches in the future.

Tests are already commercially available to help guide treatment decisions. The Oncotype DX test is a 21 gene assay performed in paraffin-embedded formalin fixed tissue and is now well established for use in ER+, LN- tumors. A recurrence score (RS) based on the expression of the 21 genes can be used to predict low, intermediate or high 10-year risk of relapse-free survival. Goldstein, et al. explored the prognostic utility ofOncotype DX in patients enrolled on Intergroup trial E2197, a study that included patients with 0-3 positive lymph nodes and either ER+ or ER- disease. The primary study objective was to test AC vs. doxorubicin and docetaxel (AT) head-to-head with tamoxifen given after chemotherapy if the patient was ER+. As per previous datasets in the hormone receptor-positive group, 46% of patients in this trial had a low RS, 30% of patients had an intermediate RS and 24% had a high recurrence score. In contrast, 99% of patient with ER- tumors had a high recurrent score. Therefore, the test is not suitable for use in ER- patients.

However, low RS was predictive of good relapse-free survival (RFS) regardless of whether patients were node-negative or had 1-3 lymph nodes positive. Because of the design of the study, the benefit of chemotherapy in low-risk patients above and beyond that of tamoxifen could not be assessed. Evaluation of a differential benefit in the two different chemotherapy arms has not yet been performed. Still, these results already suggest that we may be able to expand the indication for this test to patients with lymph node-positive disease.

While it is agreed that the Oncotype DX test stratifies patients into prognostic groups, is it better than standard prognostic criteria and how do the results impact oncologist decision making? These questions were the subject of two reports at ASCO.

Kumal et al reviewed the Mayo Clinic database of ER+, LN- patients who had undergone Oncotype DX testing. Academic oncologists were presented patient data without knowing the RS. They were able to identify high RS patients easily, but over 1/3 of the time, predicted low RS were determined to be intermediate or predicted intermediate RS turned out to be low. Recommendations for therapy changed 18% of the time based on the recurrence score.

Lo, et al. evaluated oncologist treatment recommendations pre and post an Oncotype DX assay at 3 academic and one community based center. The assay recurrence score led to a change in treatment recommendations 31.5% of the time, 22.5% chemotherapy plus hormonal therapy to hormonal alone, 3.4% hormone to chemo+ hormone, and 5.6% to equipoise from chemo+ hormone or hormone alone. These studies confirm the clinical usefulness of the Oncotype DX test over standard pathologic criteria alone.

Endocrine Therapy

No major new adjuvant endocrine trials were presented at ASCO this year. Nonetheless, there continues to be important data generated by further evaluation of the datasets that compared aromatase inhibitors (AIs) to tamoxifen in a variety of strategies as hormonal therapy for the adjuvant treatment of breast cancer.

While it is known that AIs improve outcome compared to tamoxifen, the magnitude of benefit in both DFS and OS is less clear. Ciccarese performed a meta-analysis of ten randomized clinical trials comparing AIs to tamoxifen. Regardless of sequencing strategy for the AI (upfront, switching after 2-3 years, or extended after five years tamoxifen), the AIs provide an absolute DFS improvement of 2.3% to 3.5%. A nonsignificant trend in overall survival was reported with a significant OS benefit of 1.6% defined in the early switching strategy. Strong correlation between DFS and OS was reported overall, therefore strengthening the case that DFS is a good surrogate marker as a predictor for OS in the adjuvant hormonal therapy setting.

From the BIG-98 trial, an up-front comparison of letrozole to tamoxifen, an analysis of cardiovascular adverse events was presented. The results demonstrated an increase for all grade thromboembolic events for tamoxifen, relative risk 0.44 and grade 3-5 events, RR=0.38, p=<0.001, while all grade cardiac events were equivalent for tamoxifen and letrozole, RR=0.99. Severe (grade 3-5) cardiac events were significantly higher for letrozole, RR=1.63, p=0.0036. However, the absolute difference in severe cardiac events was small, representing only 1% of patients on trial. There was no difference in CVAs or TIAs between the groups. This information is important in determining best therapy in patients with very early, low-risk breast cancers and those considering chemo-prevention trials.

The ATAC trial was the first to demonstrate the benefit of an AI, in this instance anastrozole compared to tamoxifen. Initially, there appeared to be some differential benefit based on the subtypes of hormone receptor positivity. An updated subgroup analysis with centralized marker evaluation demonstrated equivalent benefit for anastrozole compared to tamoxifen in the ER+/PR+ and the ER+/PR- subgroups. Additionally, patients with HER2+ disease also appear to derive benefit from the AI compared to tamoxifen. Therefore, those studies that have looked at this issue have now failed to demonstrate a difference in AI benefit based on subgroups of ER versus PR positivity.


Studies reported at ASCO 2007 reinforced the idea that patients continue to derive increased benefit from the use of adjuvant chemotherapy and hormonal therapy. Validating and measuring targets by increasingly sophisticated diagnostic testing will allow us to better define adjuvant strategies. The goal of simultaneously improving outcome and reducing the risk of long-term toxicity from treatment appears at hand.


Sparano J, Wang M, Martino S., Phase III study of doxorubicin-cyclophosphamide followed by paclitaxel or docetaxel given every 3 weeks or weekly in operable breast cancer: Results of Intergroup Trial E1199. Proceeding from the American Society of Clinical Oncology Conference .2007. Chicago, IL. Abstract # 516.

Loesch D, Greco F, O’Shaughnessy J., et al. A randomized, multicenter Phase III trial comparing doxorubicin + cyclophosphamide followed by paclitaxel or doxorubicin + paclitaxel followed by weekly paclitaxel as adjuvant therapy for high-risk breast cancer. Proceeding from the American Society of Clinical Oncology. 2007. Chicago, IL. Abstract # 517.

Andre F, Broglio K, Roche H., et al. Estrogen receptor expression and efficacy of docetaxel in early breast cancer: A pooled analysis of 3,490 patients included in two randomized trials. Proceeding from the American Society of Clinical Oncology Conference .2007. Chicago, IL. Abstract # 537.

Perez E, Romond E, Suman V., et al. Updated results of the combined analysis of NCCTG N9831 and NSABP B-31 adjuvant chemotherapy with/without trastuzumab in patients with HER2-positive breast cancer. Proceeding from the American Society of Clinical Oncology Conference .2007. Chicago, IL. Abstract # 512.

Paik S, Kim C, Jeong J., et al. Benefit from adjuvant trastuzumab may not be confined to patients with IHC 3+ and /or FISH-positive tumors: Central testing results from NSABP B-31. Proceeding from the American Society of Clinical Oncology Conference .2007. Chicago, IL. Abstract # 511.

Rastogi P, Jeong J, Geyer C., Five year update of cardiac dysfunction on NSABP B-31, a randomized trial of sequential doxorubicin/cyclophosphamide (AC)? Paclitaxel (T) vs. AC? T with trastuzumab (H). Proceeding from the American Society of Clinical Oncology Conference .2007. Chicago, IL. LBA # 513.

Dumontet M, Reed J, Krajewska M, et al. BCIRG 001 molecular analysis: Identification of prognostic factors in patients(pts) received adjuvant therapy for node-positive (N+)breast cancer (BC). Proceeding from the American Society of Clinical Oncology Conference .2007. Chicago, IL. Abstract # 525.

Press M, Sauter G, Buyse M., et al. Alteration of topoisomerase II-alpha gene in human breast cancer and its association with responsiveness to anthracycline-based chemotherapy. Proceeding from the American Society of Clinical Oncology Conference .2007. Chicago, IL. Abstract # 524.

Goldstein L, Gray R, Childs B, et al. Prognostic Utility of the 21-Gene Assay in Hormone Receptor (HR) Positive Operable Breast Cancer and 0-3 Positive Axillary Nodes Treated with Adjuvant Chemohormonal Therapy (CHT): An Analysis of Intergroup Trial E2197. Proceeding from the American Society of Clinical Oncology Conference .2007. Chicago, IL. Abstract # 526.

Kamal A, Loprinzi C, Reynolds, C How well do standard prognostic criteria predict oncotype DX(ODX) scores. Proceeding from the American Society of Clinical Oncology Conference .2007. Chicago, IL. Abstract # 576.

Lo S, Norton J, Mumby P, et al. Prospective multicenter study of the impact of the 21-gene recurrence score (RS) assay on medical oncologist (MO) and patient (pt) adjuvant breast cancer (BC) treatment options. Proceeding from the American Society of Clinical Oncology Conference .2007. Chicago, IL. Abstract # 577.

Ciccarese M, Bria E, Cuppone F., et al. Disease-free survival (DFS) as surrogate end point for overall survival (OS) in adjuvant aromatase inhibitors (AIs) trials for breast cancer (BC): Meta-analysis of 10 randomized clinical trials (RCTs) exploring the magnitude of the benefit. Proceeding from the American Society of Clinical Oncology Conference .2007. Chicago, IL. Abstract # 539.

Coates A, Mouridsen H, Sun Z., et al. Cardiovascular adverse events during adjuvant endocrine therapy for early breast cancer using letrozole or tamoxifen: Updated safety analysis of trial BIG 1-98. Proceeding from the American Society of Clinical Oncology Conference .2007. Chicago, IL. Abstract # 521.

Rasmussen B, Regan M, Lykkesfeldt A., et al. Central assessment of ER, PgR and HER2 in BIG 1-98 evaluating letrozole (L) compared to tamoxifen (T) as initial adjuvant endocrine therapy for postmenopausal women with hormone receptor-positive breast cancer. Proceeding from the American Society of Clinical Oncology Conference .2007. Chicago, IL. Abstract # 538.

Copyright © 2018 CancerConnect. All Rights Reserved.